Skip to main content

Maintenance · UK

Public charging reliability method

Treat the statutory 99% figure as an annual rapid-network metric, not a visit guarantee. It is operator-reported over a calendar year, rather than a promise about one charger. Raw status observations, controlled measurements and modelled scores remain separate from that annual report. Because the current E2 build has no comparable licensed observations, EV Evidence publishes no operator score or ranking.

The National Archives: The Public Charge Point Regulations 2023, Regulation 7

Reviewed

Key facts and evidence

For UK public charging reliability reviewed on 13 August 2026, the statutory 99% figure is an operator-reported calendar-year average for a qualifying rapid network, not a charger, site, month or future-visit guarantee. Observed point status, measured test results and modelled scores remain separate. E2 currently has no comparable licensed observations, so no operator score or ranking is published.

The National Archives: The Public Charge Point Regulations 2023, Regulation 7

How EV Evidence separates operator annual reports, observed charger status, controlled measurements and modelled reliability outputs.

Operator-reported annual metric

For an operator that is not a micro business, the Regulations set an average 99% calendar-year reliability requirement for its rapid public network, defined at 50 kW or above, with the applicable treatment of restricted hours. EV Evidence labels the published operator figure operator-reported and preserves its year, network scope, formula and exemptions.

Observed status

An observed status is a timestamped record for a defined EVSE or charger from an admitted feed. It retains source latency, identifier, location and observation time. One available or unavailable snapshot does not establish annual compliance, and missing status remains unknown rather than unavailable or zero.

Measured tests

A measured result requires a stated test protocol, sampling frame, date, equipment or transaction definition, failures and denominator. Mystery-shop or field-test results are not relabelled as operator annual reports and are not generalised beyond the sampled points and period.

Modelled output

A modelled score can be produced only from comparable licensed observations with complete metric definitions, periods and sample gates. Its method, components and missing-value handling must remain visible. It is independent analysis, not a regulatory finding or source endorsement.

Current E2 boundary

The current E2 charging scorecard has an empty operator array and the reason no comparable licensed observations. Consequently the score, ranking, stars and comparative recommendation remain absent. The method is documented, but an unpopulated denominator is reported as insufficient evidence rather than 0%.

What to do next

Use reliability evidence for planning only and follow the operator's live instructions and the relevant safety page at the chargepoint.

Immediate action: Check the evidence label, period, network scope, sample and observation time before interpreting a figure.

  • Stop relying on a figure if its evidence type, calendar period, denominator, rapid-network scope or freshness is missing.

Professional hand-off

  • Report a point fault to the chargepoint operator; direct a regulatory scope or compliance concern to the Office for Product Safety and Standards.

Questions this record can answer

What does the sourced evidence say about operator-reported annual metric?

For an operator that is not a micro business, the Regulations set an average 99% calendar-year reliability requirement for its rapid public network, defined at 50 kW or above, with the applicable treatment of restricted hours. EV Evidence labels the published operator figure operator-reported and preserves its year, network scope, formula and exemptions.

What does the sourced evidence say about observed status?

An observed status is a timestamped record for a defined EVSE or charger from an admitted feed. It retains source latency, identifier, location and observation time. One available or unavailable snapshot does not establish annual compliance, and missing status remains unknown rather than unavailable or zero.

What does the sourced evidence say about measured tests?

A measured result requires a stated test protocol, sampling frame, date, equipment or transaction definition, failures and denominator. Mystery-shop or field-test results are not relabelled as operator annual reports and are not generalised beyond the sampled points and period.

Sources